duckdb 0.9.1-dev43.0 → 0.9.1-dev69.0

package/package.json

@@ -2,7 +2,7 @@
   "name": "duckdb",
   "main": "./lib/duckdb.js",
   "types": "./lib/duckdb.d.ts",
-  "version": "0.9.1-dev43.0",
+  "version": "0.9.1-dev69.0",
   "description": "DuckDB node.js API",
   "gypfile": true,
   "dependencies": {

package/src/duckdb/src/execution/operator/csv_scanner/sniffer/csv_sniffer.cpp

@@ -22,39 +22,37 @@ CSVSniffer::CSVSniffer(CSVReaderOptions &options_p, shared_ptr<CSVBufferManager>
 	}
 }
 
+void CSVSniffer::SetResultOptions() {
+	options.dialect_options = best_candidate->dialect_options;
+	options.dialect_options.new_line = best_candidate->dialect_options.new_line;
+	options.has_header = best_candidate->dialect_options.header;
+	options.skip_rows_set = options.dialect_options.skip_rows > 0;
+	if (options.has_header) {
+		options.dialect_options.true_start = best_start_with_header;
+	} else {
+		options.dialect_options.true_start = best_start_without_header;
+	}
+}
+
 SnifferResult CSVSniffer::SniffCSV() {
 	// 1. Dialect Detection
 	DetectDialect();
-	if (explicit_set_columns) {
-		if (!candidates.empty()) {
-			options.dialect_options.state_machine_options = candidates[0]->dialect_options.state_machine_options;
-			options.dialect_options.new_line = candidates[0]->dialect_options.new_line;
-		}
-		// We do not need to run type and header detection as these were defined by the user
-		return SnifferResult(detected_types, names);
-	}
 	// 2. Type Detection
 	DetectTypes();
 	// 3. Header Detection
 	DetectHeader();
+	if (explicit_set_columns) {
+		SetResultOptions();
+		// We do not need to run type refinement, since the types have been given by the user
+		return SnifferResult({}, {});
+	}
 	D_ASSERT(best_sql_types_candidates_per_column_idx.size() == names.size());
 	// 4. Type Replacement
 	ReplaceTypes();
 	// 5. Type Refinement
 	RefineTypes();
-	// We are done, construct and return the result.
-
-	// Set the CSV Options in the reference
-	options.dialect_options = best_candidate->dialect_options;
-	options.has_header = best_candidate->dialect_options.header;
-	options.skip_rows_set = options.dialect_options.skip_rows > 0;
-	if (options.has_header) {
-		options.dialect_options.true_start = best_start_with_header;
-	} else {
-		options.dialect_options.true_start = best_start_without_header;
-	}
-
-	// Return the types and names
+	// We are done, Set the CSV Options in the reference. Construct and return the result.
+	SetResultOptions();
 	return SnifferResult(detected_types, names);
 }
 

package/src/duckdb/src/function/table/read_csv.cpp

@@ -38,7 +38,7 @@ void ReadCSVData::FinalizeRead(ClientContext &context) {
 	auto number_of_threads = TaskScheduler::GetScheduler(context).NumberOfThreads();
 	//! If we have many csv files, we run single-threaded on each file and parallelize on the number of files
 	bool many_csv_files = files.size() > 1 && int64_t(files.size() * 2) >= number_of_threads;
-	if (options.parallel_mode != ParallelMode::PARALLEL && many_csv_files) {
+	if (options.parallel_mode != ParallelMode::PARALLEL && (many_csv_files || number_of_threads == 1)) {
 		single_threaded = true;
 	}
 	if (options.parallel_mode == ParallelMode::SINGLE_THREADED || not_supported_options ||

package/src/duckdb/src/function/table/version/pragma_version.cpp

@@ -1,8 +1,8 @@
 #ifndef DUCKDB_VERSION
-#define DUCKDB_VERSION "v0.9.1-dev43"
+#define DUCKDB_VERSION "v0.9.1-dev69"
 #endif
 #ifndef DUCKDB_SOURCE_ID
-#define DUCKDB_SOURCE_ID "709ea292d3"
+#define DUCKDB_SOURCE_ID "5ba1abd81a"
 #endif
 #include "duckdb/function/table/system_functions.hpp"
 #include "duckdb/main/database.hpp"

package/src/duckdb/src/include/duckdb/execution/operator/scan/csv/csv_sniffer.hpp

@@ -50,6 +50,8 @@ private:
 	CSVReaderOptions &options;
 	//! Buffer being used on sniffer
 	shared_ptr<CSVBufferManager> buffer_manager;
+	//! Sets the result options
+	void SetResultOptions();
 
 	//! ------------------------------------------------------//
 	//! ----------------- Dialect Detection ----------------- //

package/src/duckdb/src/optimizer/common_aggregate_optimizer.cpp

@@ -38,8 +38,8 @@ void CommonAggregateOptimizer::ExtractCommonAggregates(LogicalAggregate &aggr) {
 			// aggregate does not exist yet: add it to the map
 			aggregate_remap[*aggr.expressions[i]] = i;
 			if (i != original_index) {
-				// this aggregate is not erased, however an agregate BEFORE it has been erased
-				// so we need to remap this aggregaet
+				// this aggregate is not erased, however an aggregate BEFORE it has been erased
+				// so we need to remap this aggregate
 				ColumnBinding original_binding(aggr.aggregate_index, original_index);
 				ColumnBinding new_binding(aggr.aggregate_index, i);
 				aggregate_map[original_binding] = new_binding;